Dushyanth <dushyanth@xxxxxxxxx> writes: >> Tom Lane <tgl <at> sss.pgh.pa.us> writes: >> Then the only other possibility is that autovacuum is being launched to >> prevent XID wraparound. Are there any tables in that database with >> particularly old relfrozenxid? > Below are the unique age(relfrozenxid) values that i see from the above query > 140835139 > 150945753 > 185741480 Hmph. Does pg_database.datfrozenxid agree with that? regards, tom lane